Experiments with aerials and positioning

Hello,

I was hoping to get my new Flightaware 26" aerial mounted on a pole on my roof this weekend but none of the companies I e-mailed were able to come out this weekend.

I used to have a small stick aerial that plugged into the blue Flightaware Pro Stick Plus, this gave me mediocre results of some 600 aircraft a day and around 40K positions here in Bristol (Site 30123). As I live between the UK and the USA I also have a NYC site (30998) which performs significantly better with the same configuration in a much more built up area in terms of buildings. It would on a daily basis report some 170K positions and around 2400 aircraft. I’ve been very surprised at the performance of this site considering the high level of building clutter near the aerial. I recently (5 April 2019) upgraded the aerial in NYC to the Flightaware 26" aerial but so far the results do not seem that different to when using the small stick aerial. This site also has the Flightaware 1090Mhz filter, I’ve not really experimented with that other than to install it. I’ve not tried it without the filter.

At my Bristol site, I decided to experiment with the new Flightaware 26" aerial. Replacing the stick aerial with this gave immediate results even with the huge 5M cable which was the smallest size I could readily get off the shelf. With such a huge amount of cable I wondered what it would be like to move the aerial around upstairs. First I positioned it in another window, which made the performance worse. Then I tried propping it up against an outside wall of the house that faced West. Pretty much instantly the receiver went from 90 messages a second to over 400 a second.

The next day I decided to get a little more adventurous and I went to Wickes and purchased a step ladder to see what was in my loft and whether I could position and aerial up there. I found a wooden beam nearest to the outside wall and cable tied the 26" Flightaware aerial to that as high up as I could get it. The number of messages per second went from 400 to around 600 and I could also see planes as far north as Sheffield and Manchester, occasionally planes as far south as northern France. The first full day in this configuration bagged me 1589 aircraft reported and 169K positions which I am quite pleased with.

I have ordered a N Male to SMA Male adapter i.e. no cable at all and that will arrive tomorrow. I plan to plug the Flightaware Pro Stick Plus directly into the aerial and then use a USB extender to plug the receiver into the Raspberry PI. That way I don’t need to cable tie the rPI to the same wooden beam. I hope this will significantly reduce the loss in the 5M cable and improve results even further. I’d like to get to at least 2K aircraft reported from my Bristol site.

1 Like

If you use the FA antenna, you might have to lower gain to get better results.

Even though the blue FA stick has a filter, It might help to add a second one.

There are two FA filters, a light blue version for the US, and a dark blue version for Europe, due to the different mobile phone frequencies.

Some people had issues with inference from usb cables, so it might be actually better to have a short antenna cable from antenna to the dongle + Pi.

I just experimented a little with the Pro Stick Plus gain. It is currently set to -10 which I guess is automatic gain from what I read. I set it to 50, 45, 40, 35 and 30 and restarted each time. It didn’t seem to make a huge difference. I might attempt a more scientific approach later today! I have uploaded a screenshot of what my receiver is currently seeing, there does appear to be some chunks of the country where I don’t see any planes. If I look at Plane Finder on my phone, it seems that there are a lot of flights nearby that I cannot “see”.

That value translates into a gain value of around 55, higher than you can set manually. AGC is meant to work with a constant signal like TV the dongle tunes itself to, ADS-B are many short messages, so the dongles ends up using the highest possible gain.

Generally, I found the FA antenna is much better than those basic antennas, but also it is more sensitive to its location and surroundings

If you click on a plane, Skyview will show the RSSI value of the signals, they can also give an indication whether the dongle is overloaded. There a quite a few threads here discussing gain settings, so just look around.

Take a look at heywhatsthat.com, it allows you to find out how you are limited by terrain. You can actually import the results to create additional distance rings.

Light blue or dark blue? The dark blue is a new design for Europe which better removes the 900 MHz GSM band.
If you go to the trouble of paying for the aerial to be installed you might want to consider the uputronics LNA which is probably easiest to use and costs 40 pounds i believe. (It has a very good builtin filter, together with the prostick you would need to turn the gain way down but it should work)

Depending on the brand of extender such devices can cause the receiver to hang or MLAT to stop working. Wouldn’t recommend them.

Thanks for the suggestion about the USB extender, I’ve tried it with and without and I can’t see any difference but I will bear in mind what you said.

As for the filter, it’s blue and I purchased it in the USA so I’m hoping it’s the one for the USA market.

Yeah but it’s not quite the best filter for Great Britain.
The US has different frequencies in use for mobile phones.
So getting either the dark blue filter or even better the uputronics LNA might help your reception.

I have two sites, one in the UK (Bristol) and the other one in NYC where I also live. I purchased the blue filter for the USA, I do not have one in the UK.

I’ve looked at the Uputronics amp, it looks interesting and I may experiment with that in due course.

Well not even the best antenna or amplifier or filter can help with buildings that are in the way.
But in NYC there are many airplanes flying high over the horizon.

To get similar numbers in Bristol you need to have much better range than in NYC.
Also there is more clutter on 1090 MHz in Europe because the radars interrogate at shorter intervals compared to the US.
Also mobile frequencies are closer in the frequency spectrum than the US.

So you really can’t compare those two stations.

An update on my experiment.

I now have the Pro Stick Plus plugged directly into the 26" Flightaware aerial in two configurations. First with the USB extender, second without the USB extender. So far it doesn’t appear to make any different to the number of positions or aircraft.

When I first plugged it all together the gain was still on 50 and I’ve experimented with tweaking that number downwards. At 50 it “seemed” to have fewer planes, not a lot of science behind that I will admit. I looked at an article about adjusting gain and ran the suggested awk and get this:

Percentage of strong messages: 6.517

I may adjust the gain down slightly to get to the suggested 5% number but then again I am quite close to Bristol airport.

Since making the change (plugging the receiver directly into the aerial) I notice that now I see groups of planes closely together which I have not noticed before. For instance, I see larger groupings of transatlantic flights in the morning headed to LHR or FRA.

I’ll run this configuration for a few days and wait for it to settle down a little before drawing any more conclusions. I think the only further significant modification would to have the aerial mounted on the roof on top of a pole.

Installing graphs can be useful. While traffic does change it gives some indication.

GitHub - wiedehopf/graphs1090: Graphs for readsb / dump1090-fa / dump1090 (based on dump1090-tools by mutability)

Your Flightaware stats page shows quite a significant improvement for Saturday:

https://flightaware.com/adsb/stats/user/jnyc

Saturday is when I mounted the aerial in the loft, there was considerable improvement with this change.

I have downloaded graphs but it seems to be stuck on the download and install:

Setting up python2.7 (2.7.13-2+deb9u3) …
Setting up libpython2.7:armhf (2.7.13-2+deb9u3) …
Processing triggers for libc-bin (2.24-11+deb9u3) …

It’s been stuck on the last line for around 10 minutes now, I’m not sure if that is normal or not.

That doesn’t normally happen.

Press Ctrl-c (maybe more than once until you have the prompt again).
Then try again.

Thanks, it has now installed but the graphs are empty so I guess I need to leave it for a while. I will report back tonight when I am back from work!

Here are a few of the graphs:




Hello,

This is what my setup looks like after a full 24 hours: